Avoid empty file creation - perl

Question

I want to split the large file in to small files by splitting at the specific line with the help of regex. Any help?
My code doing the job but it also creating a empty file.
#!/usr/local/lib/perl/5.14.2
open( INFILE, 'test.txt' );
#lines = <INFILE>;
$file = "outfile";
for ( $j = 0; $j <= $#lines; $j++ ) {
open( OUTFILE, ">", $file . $j );
$file_name = $file . $j;
#print "file is $file_name\n";
$i = 0;
while (#lines) {
$_ = shift #lines;
chomp;
$i++;
if ( $_ =~ /^###\s*(.*)\s*###/ && $i > 1 ) {
unshift #lines, "$_\n";
print "$filename\n";
last;
}
print OUTFILE "$_\n";
}
close(OUTFILE);
}
close(INFILE);
My input file contains :
-------------
### abcd hdkjfkdj ####
body 1 dsjklsjdfskl
### zyz fhid ###
abcdksdsd djnfkldsfmnsldk ;lkjfkl
---------------------------
it is creating 3 outfiles called outfile0,outfile1,outfile2. but outfile0 is empty I want to avoid this.

Answer 1

The way to fix it is to open the file only in response to the line being found. Your program will open a new file regardless and that's why it has an empty output file
Here is a rewrite that works. I've also removed the temporary #lines array
#!/usr/bin/perl
#
use warnings;
use strict;
open(my $file,"<", "test.txt") || die $!;
my $counter=1;
my $out;
while(<$file>) {
if (/###\s*(.*)\s*###/) {
open($out, ">", "outfile$counter") || warn "outfile$counter $!";
$counter++;
}
print $out $_ if $out;
}

Answer 2

If you want to use the material between the ### blocks as file titles, you can set the file name when you're doing the pattern match on the lines with ### blocks.
#!/usr/bin/perl
use strict;
use warnings;
open my $fh, '<', 'my_file.txt' or die "Could not open file: $!";
# initialise a variable that will hold the output file handle
my $out;
while (<$fh>) {
# capture the title between the # signs
if (/##+ (.*?) ##+/) {
open $out, '>', $1.".txt" or die "Could not create file $1.txt: $!";
}
elsif ($out) {
print $out $_;
}
else {
# if $out is not set, we haven't yet encountered a title block
warn "Error: line found with no title block: $_";
}
}
Sample input:
Text files containing their own name
### questions-1 ####
Why are a motorcycle's front brakes more effective than back?
Is it possible to make a gradient follow a path in Illustrator?
Text files containing their own name
### questions-2 ###
Why does Yoda mourn the Jedi after order 66 is executed?
what are the standard gui elements called?
Flybe just cancelled my return flight. Will they refund that part of the trip?
### questions-3 ###
Merge two arrays of ElementModels?
Is this set open or closed?
Output: three files, questions-1.txt, questions-2.txt, questions-3.txt, containing the appropriate lines. e.g. questions-1.txt:
Why are a motorcycle's front brakes more effective than back?
Is it possible to make a gradient follow a path in Illustrator?
Text files containing their own name
You haven't stated whether you want the ### lines in the output or not, so I've left them off.
Depending on what OS you're on and what your potential file names contain, you may want to filter them and replace special characters with an underscore (or just remove the special characters).